Member Junction
    Preparing search index...

    Interface ClusterResult<TMetadata>

    Result of clustering operations

    interface ClusterResult<TMetadata = Record<string, unknown>> {
        centroids?: Map<number, number[]>;
        clusters: Map<number, string[]>;
        metadata?: {
            inertia?: number;
            iterations?: number;
            metric: DistanceMetric;
            silhouetteScore?: number;
        };
        outliers?: string[];
    }

    Type Parameters

    • TMetadata = Record<string, unknown>
    Index

    Properties

    centroids? clusters metadata? outliers?

    Properties

    centroids?: Map<number, number[]>

    Map of cluster ID to centroid vector (for K-means)

    clusters: Map<number, string[]>

    Map of cluster ID to array of vector keys in that cluster

    metadata?: {
        inertia?: number;
        iterations?: number;
        metric: DistanceMetric;
        silhouetteScore?: number;
    }

    Metadata about the clustering operation

    Type Declaration

    • Optionalinertia?: number

      Sum of squared distances to centroids (for K-means)

    • Optionaliterations?: number

      Number of iterations until convergence

    • metric: DistanceMetric

      Distance metric used

    • OptionalsilhouetteScore?: number

      Silhouette score (-1 to 1, higher is better)

    outliers?: string[]

    Array of vector keys identified as outliers (for DBSCAN)

    Settings

    Member Visibility

    On This Page

    Properties